“It was probably that guy?”
– The functions of reconstructive speech acts in investigative
training interviews
Winnie Collin Roskilde University wcollin@ruc.dk
Abstract*
Th is paper explores the pragmatic and interactional functions of reconstructive speech acts in mock police interviews, based on a model of argumentative dialogue. Th e aim of the paper is to illustrate how the reconstructions apparently contribute to both the interaction between the police offi cer and the mock suspect in the interview activity and to the interaction in the training activity, i.e. between the participants attending the training course. Drawing on functional pragmatics and grammar, the analysis seeks to examine how reconstructions on the one hand function in the socially non-cooperative interaction in the mock interview, questioning the truth value of propositions and trustworthiness of the suspect, and, on the other hand seem to fulfi l a supportive purpose in the training activity.
Keywords: Investigative interviewing, training activities, reconstructive speech acts, pragmatics, semantics, questions, forensic linguistics
1. Introduction
Taking its point of departure in speech act theory and functional grammar, this study focuses on reconstructive speech acts’ (see section 3) questioning and evaluating functions across diff erent activity types embedded in simulated police interviews.
Police interviews with witnesses and suspects are one of the main sources for information during criminal investigation. In Denmark, such investigative
interviewing is inspired by the British PEACE-model1 and principles from cognitive interviewing (Fischer and Geiselman 2010) and follows four stages:
Preparation, free account, questioning and closure (Bitsch and Modin 2016:82, Gibbons 2003:145-146). Th e excerpts in this paper are from the stage of questioning, in which the interviewer “in the following order (…) will tend to use a lower question type if a higher one does not work (a) open questions (b) either/Or multiple choice questions (c) Specifi c but not leading questions (d) leading questions” (Gibbons 2003:145-146). Th e aim of the investigative interview is to clarify the circumstances of the event under investigation and as objectively and accurately as possible, elicit “correct information in an unbiased and open-minded manner compared to a question-answer approach” (Hviid 2017:30).
Since 2005, the Danish police has conducted investigative interviews based on the Danish interview model. Furthermore, interview training is a substantial part of the education and ongoing training of police offi cers. Part of the training is conducted as simulated investigative interviews with police offi cers playing the roles of witnesses, suspects, and interviewer. How the participants of such training interviews handle the shift ing communicative roles, and how they make sense of the contributions to this complex kind of interaction, has not yet been studied in a Danish context. Similar aspects of this kind of complex communication have been the focus of other studies, though, among others Linell and Th unqvist’s study of simulated job interviews. Linell and Th unqvist (2003:413) describe how participants involved in communication that involves teaching are oft en “inquiring about what is going on, and rehearsing, evaluating, and making meta-comments on activities and happenings”. Th us, Linell and Th unqvist (2003:431) notes, the participants move in and out of framings2 and “build their utterances to fi t a specifi c activity context, or to fi t several such contexts at the same time, and they enact role identities associated with these activities”.
In the data of the present study, the participants also seem to orient to diff erent framings and the activity types (Levinson 1992) embedded in the interview training-situation. Most of the time, though, evaluation and meta- comments are phrased as utterances that can count as contributions to both the training activity and the interview activity. Th e excerpts presented in this paper are illustrative of how interlocutors in complex communicative activities, with distinctly diff erent inferential frames to derive inference from, manage to navigate and switch between activities. Reconstructive speech acts seem to be a helpful tool for this. Th us, the analysis serves to illustrate how diff erent activity types make certain inferences possible and acceptable, and to demonstrate how a model of argumentative dialogue holds useful concepts for analysing the negotiation of meaning across activities.
In authentic as well as mock investigative interviews, the reconstructive speech acts seem to serve the purpose of open questions well as they invite the interviewee to provide a more or less free account. However, a police
interview is an activity type where “police offi cers tend to use questions to glean information, query, accuse, etc., with the ultimate aim of testing the veracity of a witness’s or suspect’s account of events (Archer 2020, this issue, see also Vrij 2008). Th at is, an activity type with a built-in confl ict or divergent points of view. On that basis, Nielsen’s model of argumentative dialogue (Nielsen 2005) establishes a useful starting point for the analysis of the interaction in training interviews, in particular the notions of reconstruction and criticism.
Among others, Nielsen’s model draws on Searle’s (1976) speech act theory and Grice’s (1975) cooperative principle that, accordingly, form the theoretical and analytical framework of the present study. In addition, the analysis of the contextual pragmatic functions of the reconstructions draws on Danish Functional Grammar (Hansen and Heltoft 2019) and studies of epistemic stance in Danish spoken language (Mortensen 2012) and in courtroom interaction (Mortensen and Mortensen 2017).
An evaluating function of reconstructions is the focus of this paper, that is, reconstructions that function as questions while at the same time evaluate certain aspects of the interviewee’s3 utterances – apparently across the activities embedded in the training interview. My labelling of the reconstructions as questions is based on semantic and contextual interpretations (see also section 4.1), but since the aim of this paper is not to determine whether the reconstructions are to be considered questions or not, I will refrain from discussing ways to identify and classify questions.
2. Th e training interview activity
Th e interaction analysed in this study takes place as a mock police interview during a training course for police offi cers. Th us, there are, at least, two diff erent activity types (Levinson 1992), the interview activity and the training activity, embedded in the overarching and complex activity, the training interview.
Th e participants in the training interview probably know each other, at least from attending the same training course for some days. Within the context of the training activity, the relation is symmetric, and the goals for the participants are probably more or less the same; that is, training and improving their interviewing skills and performing well in order to make the mock interview proceed the best way possible. Consequently, the participants have a common interest in keeping the mock interview on the right track and play their roles as they are instructed to do.
In the context of the interview activity, the interviewees follow their instructions about what they are supposed to tell the interviewer – and what they do not want to tell. Th e participants try to simulate an asymmetric relation, but obviously, the conditions for the interaction in the interview activity are not the same as the conditions for an authentic police interview, neither regarding the relation between the participants or possible confl icting goals and interests.
As already mentioned, the training interview holds two diff erent activity types. On the one hand, these activities constitute diff erent framings for the interlocutors to act within, and on the other hand, they intertwine. Activity types are ”a fuzzy category whose focal members are goal-defi ned, socially constituted, bounded, events with constraints on participants, setting, and so on, but above all on the kinds of allowable contributions” (Levinson 1992:69).
Hence, the notion of activity types is a useful tool for distinguishing and analysing the interaction and the interlocutors’ speech acts in the training interview.
Activities hold distinctive structural elements that are ”rationally and functionally adapted to the point or goal of the activity in question, that is the function or functions that members of the society see the activity as having”
(Levinson 1992:71). Th us, the activity type constrains which roles it is possible for the interlocutors to have, and how they are able to contribute to the interaction, linguistically as well. Speech can be a more or less integrated part of an activity, and the organisation of the activity both constrains “what will count as an allowable contribution to each activity”, and constitutes certain activity-specifi c rules of inference that “help to determine how what one says will be ”taken” – that is, what kinds of inferences will be made from what is said”
(Levinson 1992:97). As regards rules of inference, the training activity and the interview activity constitute quite diff erent framings for the participants to derive meaning from. In addition, the interview activity simulates a relation between participants with assumed confl icting goals and assumed lack of social cooperation4 (Archer 2005:58). Even so, Levinson (1992:77) notes,
”there could be some quite interesting relations between Grice’s maxims and diff erent kinds of activities, of a sort where some of the maxims are selectively relaxed to varying degrees in activities of specifi c types”. Levinson (1992:78) argues that even in conversations that are not inherently cooperative, i.e. in the social sense of the word, it is possible to talk about relations between specifi c activity types and Grice’s maxims, if one ”accept[s] Grice’s maxims as specifi cations of some basic unmarked communication context, deviations from which, however common, are seen as special or marked”.
3. Th eoretical background
Th e notion of reconstruction employed in this study, is inspired by the model of argumentative dialogue put forward by Nielsen (2005, 2010), in which the counter argumentative speech act types reconstruction and criticism are discussed in detail. Th e model is rooted in the pragma-dialectical school of argumentation studies (e.g. van Eemeren and Grootendorst 1984, van Eemeren and Grootendorst 1992, Grootendorst et al. 1996, Nielsen 2005) that combines “the pragmatic focus on language as social action with the dialectical focus of rule-based confl ict resolution” (Nielsen 2005:26). In the pragma-dialectical approach, arguments are identifi ed as external, functional, social and dialectical. Consequently, studying argumentation builds on four
corresponding methodological premises that “are the basis for integrating the descriptive dimension of argumentation as an actual phenomenon of argumentative discourse with the normative dimension of argumentation as an essential ingredient of critical discussion” (Grootendorst et al. 1996:276).
Briefl y outlined, the methodological premise of functionalisation points to the focus on “the function of argumentation in the verbal management of disagreement” (Grootendorst et al. 1996:278) and as persuasive acts (Nielsen 2005:27). Th e notion of externalisation covers the focus on language and what interlocutors “have expressed in discourse, whether directly or indirectly”
(Grootendorst et al. 1996:276). Considering arguments as communicative and interactional acts, psychological dispositions and motives should not be the focus of analysis. Rather, the analysis of arguments is supposed to concentrate on the “externalized-or externalizable-commitments” of the interlocutors (Grootendorst et al. 1996:277) and “distinguish what commitments follow from the discursive contributions” (Nielsen 2005:27).
Identifying arguments as social, signals that argumentation is a dialogical process involving two or more interlocutors participating in argumentative interaction, “switching the communicative roles of speaker and hearer, and the interactional roles of protagonist and antagonist” (Nielsen 2005:27). Finally, according to Nielsen (2005:28) “dialectifi cation of argumentation refers to van Eemeren and Grootendorst’s philosophical position in critical rationalism”
and the idea that argumentation is dialogical in the sense that it is not just a process of “advancing standpoints, but just as much criticising standpoints”.
Accordingly, the dialectifi cation of argumentation analysis covers a set of rules for the dialectical conduct of critical discussion.
Nielsen (2005:185) notes that his approach to argumentative dialogue
“can be approximately characterised ‘a pragma-dialectical approach’”, stressing that “Argumentation is a pragmatic phenomenon and should be studied as such. Argumentation should be evaluated in terms of its critical adequacy as a dialectical process, according to critical rationalist principles”.
However, the pragma-dialectical approach is normative while the purpose of Nielsen’s approach is descriptive, aiming to ”show that language users do in fact occasionally reconstruct unexpressed premises in rational (co-operative/
reasonable) ways” (Nielsen 2005:188).
3.1. A model of argumentative dialogue
Argumentative dialogue involves a protagonist who expresses and defends a point of view, and an antagonist who challenges and criticises the protagonist’s argument (Nielsen 2005:204). Nielsen proposes a model that illustrates these two roles and the linguistic actions typically connected to them, including the speech acts’ linguistic expressions and functions:
PROTAGONIST ANTAGONIST
Linguistic expression
Object linguistic Meta-linguistic referents and verbs
Meta-linguistic operators
Illocution Claiming entailment and soundness
Reconstruction by
Paraphrases Implicata
Criticism by
Formal refutations Factual refutations Perlocution Persuasion/
confl ict resolution
Clarifi cation Meta-persuasion
Type of argument
Argument Warranting counter argument
Counter argument
Figure 1: Model of argumentative dialogue (Nielsen 2005:204)
Th e antagonistic argumentation involves the meta-linguistic utterance reconstruction and the meta-argumentative criticism. Neither of these are
“ committed to particular direct speech act types”, but “may be phrased as assertions, questions, etc.” (Nielsen 2005:198). Th e purpose of reconstructions is to clarify the protagonist’s argumentation by reproducing what the protagonist has literally uttered or by “indicating what the protagonist meant by what he said, or what he is further committed to mean in the given situation”
( Nielsen 2005:36). Th e reconstruction forms the basis of the antagonist’s criticism, which is the core of the counter argument and criticises or evaluates the protagonist’s argument (Nielsen 2005:209).
Following Searle’s taxonomy reconstructions are ”a special type of representative speech act. But whereas the typical representative speech act aims at bringing ’the words to fi t the world’, this type aims at bringing the words to fi t some other words” (Nielsen 2005:206). Furthermore, Nielsen (2005:206) argues, the essential condition regarding the reconstruction, is more complex because “the speaker/antagonist is non-committed to the truth of the propositional content conveyed by the meta-linguistic referents, while he is committed to the truth of the meta-linguistic verbs and operators”. Th at is, the antagonist is committed to the consistency between the propositional content, or the intended meaning, and his reconstruction of this. Consequently, a reconstructive speech act “can never in itself add up to critical argument”
(Nielsen 2005:209).
Hence, Nielsen (2005:206) proposes the following felicity conditions for reconstructive speech acts, 5 stressing that ”it is crucial to operate with two dif- ferent propositional contents, the propositional content of the reconstructive speech act itself, p, and the propositional content of the reconstructed utterance, p’”.
1. Content condition (or propositional act.) 6 Any proposition p describing any proposition p’.
(This is a deviation from the representative speech act whose propositional content is simply ‘Any proposition p’. But reconstruction is meta-linguistic, so in order to count as a reconstruction, the content of the utterance should be about some other utterance, either its literal expression (paraphrase) or its implicit meaning (implicatum)).
2. Preparatory condition
2.1 Proposition p’ has been uttered (directly or indirectly) at some time prior to the utterance of proposition p.
2.2 S has evidence for p. It is not obvious to S and H that H already knows p.
2.3 Insofar as p’ has been uttered by H, H is prepared to have p’ reconstruct- ed or interpreted.
(Ad 2.1: Clearly, reconstructing some speech act presupposes that the speech act in question has actually been uttered (i.e., said, implicated, or implied) at the time of the reconstruction. 2.2 does not deviate from Searle’s account, while 2.3 simply states that it is a preparatory condition that H (if H is the protagonist) accepts to engage in meta-linguistic discourse about p’ in the given context.)
3. Sincerity condition
S believes that p accurately describes p’.
(S believes that his utterance, p, gives a charitable reconstruction of the re- constructed utterance, p’, but he does not have to believe p’.)
4. Essential condition
S takes on the obligation that p accurately describes p’.
(In producing the reconstructive utterance, S is committed to the accuracy of the fi t between meta-linguistic reconstruction and the utterance it is a recon- struction of, but S is not committed to p’ representing a state of affairs.)
Figure 2: Felicity conditions for reconstructions (Nielsen 2005:207)
As to the criticising speech act, Nielsen (2005:208-209) characterises this as ”a meta-linguistic description of the correspondence between some linguistic phenomenon and the facts, or the coherence between two or more linguistic phenomena”. Th at is, criticism as factual refutation referring to “the truth value of some particular utterance”, or as formal refutation referring to “the consistency between two or more utterances”. When performing the criticism, the speaker is only committed to his evaluation of the reconstructed propositional content. On that basis, Nielsen (2005:208) proposes the following felicity conditions for criticising speech acts, represented by q. Th e utterance that is criticised or evaluated is represented by p.
1. Content condition (or ‘propositional act’) Any proposition q evaluating any proposition p’.
(The content of the utterance should be about some other utterance(s), and it should specify S’s evaluation of that other utterance, or S’s evaluation of the relationship between those other utterances.)
2. Preparatory condition
S has evidence for q. It is not obvious to S and H that H already knows q.
(Again this condition is identical to the condition for the prototypical ‘as- sertion’, perhaps with the addition that insofar as p’ is ascribed to H, H should be prepared to accept having his utterances evaluated by S in the given context.)
3. Sincerity condition
S believes that q accurately evaluates p’.
(The speaker believes that his evaluation refers to the truth value of some utterance (factual refutation) or to the consistency between some utterances (formal refutation)), and that his evaluation is true.
4. Essential condition
S takes on the obligation that q accurately evaluates p’.
(S is committed to his utterance referring to the truth value of some utter- ance or the consistency between some utterances.)
Figure 3: Felicity conditions for criticism (Nielsen 2005:208)
As appears from the felicity conditions for reconstructions, Nielsen distinguishes between reconstructive speech acts in the form of paraphrases or implicata. A paraphrase is “an account of some proposition that has been literally said”, and “implicatum is understood as an account of some proposition which has been implicated by what has been literally said, and/or is derived as a necessary condition for the literally said to be coherent” (Nielsen 2005:206).
Nielsen (2005:175) distinguishes between implicata that are intended and non-intended by the speaker. Th e latter refers to a situation where the meaning intended by the speaker, and the meaning reconstructed by the hearer, are not identical. Nielsen suggests that “meaning is always negotiated in conversation”, and that non-intended meaning, like intended meaning, is implicatural.
Th us, he elaborates on Grice’s account of conversational implicature saying that “In accounts like Grice’s it appears that meaning is transmitted from speaker to hearer“ (Nielsen 2005:175). As opposed to this, Nielsen (2005:175) stresses that “it is the hearer, not the speaker, that actually does the inferential work” and derives the implicatum. Hence, Nielsen characterizes “the act of implicature” as a reciprocal action involving a double agency, illustrated by the fi gure below:
A proposition for the graphical representation of the agency of implicature:
Agency SPEAKER HEARER
of
Act The Implicans The Implicature The Implicatum
Figure 4: Th e agency of implicature (Nielsen 2005:176)
Nielsen (2005:176) argues that “an implicatum that was not intended by the speaker is also a result of the act of implicature, seeing that it was arrived at by the same kind of inferential procedure as traditional implicatures”. Hence, drawing on Ennis (1982, Nielsen 2005:178), Nielsen distinguishes between derived implicata as intended meaning (used assumptions), that is, “assumptions that have actually been intended by the speaker”, and commitments (needed assumptions) that are “thought to be necessary by a ‘rational judge’, i.e. some external norm for rationality”. On a theoretical level, the two kinds of implicata are diff erent, but in dialogue they are “performed in practically identical ways”, and “the practical analysis of argumentative dialogue will generally not be able to draw a distinct line between the two types of derived implicata” (Nielsen 2005:179-180).
4. Th e training interview as argumentative dialogue
Th e model of argumentative dialogue outlined above captures central features of more kinds of confl icting dialogue than argumentation per se. Indeed, the
training interview may be conceptualised through this lens. According to Nielsen (2005:136) “the fact that communication involves co-operation does not contradict the fact that argumentative dialogue is very oft en a kind of confl ict: there has to be a divergence of opinion, otherwise there cannot be argumentation in any ordinary sense of the word”. Th is point of view goes for training interviews as well.
In a police interview one might expect that the suspect ”may try to avoid committing himself to any defi nite statement of fact” (Levinson 1992:77). Th e interview in this study is not authentic, though, but during the training course, some of the participants are instructed to play the role of a suspect who is unwilling to share details with the interviewer and to lie about certain parts of the mock event. However, such individual and psychological dispositions, or lack of social cooperativeness, are not to be confused with Gricean, communicative cooperation and intentions (Archer 2005:58). Th us, when the mock suspect is lying or avoids shedding light on relevant parts of the event under investigation, he will most likely make his contributions as informative as required in the particular situation, but no more than that – that is, informative enough to count as communication and convey a communicative intention of truthfulness. As for the training activity, the participants also engage in a kind of argumentative dialogue when they reconstruct and evaluate the certainty of a co-participant’s utterances in relation to remembering and following the instructions, they have been given.
Applying the model of argumentative dialogue to the training interview, I employ the interactional roles, protagonist and antagonist, though referring to them as interviewee and interviewer respectively. As mentioned in section 3, arguments are identifi ed as social in the pragma-dialectical approach, meaning that the participants in the argumentative interaction switch interactional roles in the dialogical process. In this respect, the training interview diff ers from the model of argumentative dialogue, at least with regard to the interview activity.
In this kind of institutional interaction, “there is a signifi cant and built-in asymmetry of power and knowledge in the institutional participant roles (…) which cast the institutional member (…) as the questioner and the non- institutional member (witness, defendant, suspect, etc.) as the respondent”
(Harris 2011:282).
Hence, switching roles is not part of the interview activity, and consequently the interviewer, roughly speaking, has the interactional role of the antagonist who reconstructs and criticises the statements given by the interviewee who plays the role of a suspect, i.e. the protagonist in Nielsen’s model. Th ese are the role identities associated with and constrained by the interview activity. In terms of interactional roles, obviously, the training activity constitutes a less constraining frame for the participants to act within.
4.1. Reconstructions as questions
In Nielsen’s model, the meta-communicative reconstructive speech act serves as a clarifi cation or warning on which the meta-argumentative criticism is based.
In the excerpts presented in this paper, though, some of the reconstructions serve both a clarifying and a criticising or evaluating purpose. For that reason, I only employ the notion of reconstruction in my analysis, thus defi ning it as a speech act type that can fulfi l both the clarifying function and the criticising function.
According to Nielsen, both reconstructing and criticising utterances can be phrased as assertions, questions, etc. Based on interactional and semantic interpretations, I consider the reconstructions displayed in this paper to be functional questions in the sense that they are utterances that “eff ectively seek[s]
to elicit information, confi rmation or agreement” (Heinemann 2010:2703, see also Stivers and Enfi eld 2010). Following Mortensen’s (2020, this issue) categorisation, the reconstructions resemble declarative questions, as they “do not have morphosyntactic features that clearly make them questions”, and echo questions, as they repeat (part of) the interviewee’s utterance and elicit a reply from the interviewee as well (Mortensen 2020, this issue, see also Noh 1998).
Th e reconstructions share certain characteristics with declarative questions, seeing that declarative questions rarely function as ‘pure’ questions (Teleman, Hellberg and Andersson 1999, Mortensen 2018). Declarative questions, as well as reconstructions, are meta communicative and require an utterance prior to the declarative question (Mortensen 2018:118-119), and, equal to the essential condition of reconstructions, the speaker who utters a declarative question, is not committed to the truth of the propositional content, but to the meta linguistic operators (Mortensen 2018:124).
As the following analysis will illustrate, Nielsen’s model of argumentative dialogue makes it possible to capture how, on the one hand, the reconstructions function as questions seeking to elicit information and, on the other hand evaluate and test the interviewee’s statement and trustworthiness.
5. Method and data
Th e excerpts examined in this paper are a selection of examples from a study of the functions of reconstructions in mock police interviews. Th e original dataset consists of audio recordings of 39 mock police interviews and originates from an experimental study in the fi eld of applied criminology and police management, conducted during an interview training-course for experienced detectives (Hviid 2017). Th e mock interviews are set up as role-plays based on the same mock case. Th e participants attending the training course take turns playing the roles of the interviewer, suspect and a willing witness. Prior to playing the role of a suspect or a willing witness, the participants have read an instruction about their character, what they experienced during the event under investigation, and how much they are willing to tell the police (Hviid 2017:33). During the mock interview, the participants are allowed to consult
5.1. Th e mock case
Th e mock event under investigation is a case of violence that involves four young men7. One of them, Sedat, is a suspect and the interviewee in the excerpts in this paper. Sedat is a gang member, and by the time of the event he was hanging out with a boy from the neighbourhood, Søren. Two young men, Louai and Kim, show up. Th ey are both members of a rival gang, and soon Sedat and Louai start to argue and push each other. Someone has brought a knife, and during the scuffl e, Søren is hit and seriously injured. Kim and Louai run away, while Sedat and a witness take care of Søren.
In the mock case, two issues in particular are at stake. Firstly, who brought the knife, who stabbed Søren, and where is the knife now? Secondly, did the four young men meet by coincidence, or did Louai and Sedat set up the meeting to discuss an ongoing confl ict?
5.2. Method
An initial review of the data material showed that the interviewers characteristically repeat and paraphrase the interviewees’ statements in various ways that seem to fulfi l diff erent functions in the particular contexts.
Four interviews with mock suspects were chosen for further analysis, and based on transcriptions of these, using standard orthography, without indicating intonation, pause lengths etc., the interviews were hand-coded inductively as part of a preliminary analysis.
Based on the preliminary analysis of the four interviews, three predominant functions that the reconstructions seem to fulfi l were identifi ed; request for confi rmation, request for elaboration and evaluation. While requests for confi rmations and elaboration, roughly speaking, clarify in terms of comprehension, equalling Nielsen’s reconstructive speech act, the evaluating reconstructions question truth value and trustworthiness, more or less equalling Nielsen’s criticising speech act. Furthermore, the evaluating function diff ers from requests for confi rmation and elaboration because the interviewee either changes his or her explanation or sticks to the explanation that the interviewer evaluates or questions, thus making the dispute clear despite the interviewer’s disbelief in his or her trustworthiness.
Th e three functions formed the basis of a quantitative overview of the reconstructions’ occurrences and functions in the four interviews and, furthermore, a qualitative analysis of two of the four interviews8. Th e two interviews illustrate how the reconstructions fulfi l diff erent purposes, especially when the interaction is confl ictual, and how the reconstructions operate across the activities embedded in the training interview. Th e purpose of this paper is to illustrate sequences of interaction where the reconstructions function across the activities and in the intersection of social cooperation and non-cooperation, when the interviewer on the one hand questions the suspect’s trustworthiness, and on the other hand helps his co-participant
and supports the training activity. For this reason, the qualitative analysis is focused on reconstructions with a predominantly evaluating function.
6. Analysis
Th e excerpts9 in the analysis below are from an interview with the mock suspect Sedat. Sedat has a criminal background, and the police knows him very well. He is member of a criminal gang and is not in any way interested in talking to the police, and whenever possible he sidesteps issues that hold a potential confl ict.
In this interview, a female interviewee plays Sedat’s character. Th erefore, I refer to the interviewee as she and her, even though Sedat is male. It is very diffi cult for the interviewee to remember her instructions, which shows in various ways. Several times, she pauses during her explanation and tries to remember what to say, sometimes laughing and sometimes rereading her instruction before going on with the interview. At one point during the interview, the participants explicitly stop the interview activity and step out of the roles as suspect and detective. Th e interviewee is rather confused and tells the interviewer and the attending observer that it is very hard for her to remember the instructions and to distinguish the characters from the mock case, therefore struggling to remember which character she is actually playing, and what she is supposed to answer.
In the analysis, I will illustrate how the interviewer questions and raises doubt about the certainty of the interviewee’s statements by reconstructing the epistemic expressions in the interviewee’s propositions. Prior to the interaction in excerpt 1, the interviewee has explained that she and another suspect were arguing. Suddenly Søren, the victim, was screaming and bleeding, and the interviewee says that “something must have hit him”. Th e interviewee does not know what this “something” was, but says that it might have been a knife. In addition, she claims that she does not know who brought the knife, and now the interviewer tries to shed light on this issue.
Excerpt 110
1 IN: in that case where did the knife come from
2 IE: → well it is jo it is jo probably that guy Louaia who who had it 3 IN: → probably that guy Louaia who had it
4 IE: yes
5 IN: → yes why are you saying that it is probably him who had it can you explain a
6 → bit about that
7 IE: → well so it is jo, because when when he (p) when he started and and
8 scold me there right er 9 IN: yes
10 IE: → about, er, my family then then, er, then I jo got really angry jo and like
11 that and then I also started jo pushing him a little 12 IN: okay
13 IE: at at Louaia that so so so therefore so, er, then then all of a sudden I was afraid
14 → because then I could all of a sudden see one that that that Louaia had a knife
15 IN: → you are saying that you see a knife then 16 IE: yes
17 IN: yes
18 IE: → I actually see a knife
19 IN: → try to tell a little more about that if you try to 20 IE: yes
21 IN: recall, er, when you see the knife for the fi rst time.
In line 2 in the excerpt above, the interviewee employs the epistemic adverb nok ‘probably’, thus indicating that she is guessing or that her statement is based on inference rather than recollection or knowledge about what she saw during the event (Jacobsen 1992:15). By employing probably the interviewee
“opens the dialogical space” (Mortensen and Mortensen 2017:415, see also White 2003, Martin and White 2005), and by indicating weak epistemic support for the statement, the interviewee gives herself the opportunity to revise her statement if necessary (Mortensen 2012:67).
Th e interviewer reconstructs the proposition, emphasising probably and thereby addressing the interviewee’s uncertainty and questioning the truth value of a statement based on guesses and inferred conclusions. Considering the speech act as a declarative question, the interviewer questions the evidential basis for the statement. Th is seems to recur in the wh-question following (line 5), where the interviewer asks why the interviewee says that it is probably Louaia, thus requesting the interviewee to explain the reason (Heinemann 2010:2713) for the uncertainty rather than elaborating on the propositional content. However, the interviewer employs lidt ‘a little’ in his question, thereby downplaying the utterance and making it possible for the interviewee to revise her explanation – which she does (line 7) aft er she has consulted the written instruction for the mock case and her role as the suspect. Based on this and the following revised statement, it seems plausible
that the weak epistemic support, marked by probably (line 2), derives from the participant’s trouble remembering the instructions for the mock case and her role as the suspect Sedat in particular. Th us, the training activity and the mock interview activity intertwine, and from the perspective of the interviewee, the opening of the dialogical space might be a way of showing her uncertainty about the instructions, thereby leaving herself with the opportunity to revise her statement if it is questioned or challenged by the interviewer.
In line 7-15, the interviewee changes her explanation about the knife, now stating that Louai did have a knife. Four times in this sequence (line 7, 10, 11), she employs the Danish dialogical particle jo ‘as you know’. Th ereby, the interviewee indicates mutual agreement about her statement, as jo holds the polyphonic instruction that the speaker has a point of view that she expects the interviewer to agree upon (Hansen and Heltoft 2019:1050). Th e interviewee ends her revised explanation by saying that she faktisk ‘actually’ sees a knife.
By employing actually she indicates and recognises that an opposing point of view is possible (in this case her own inferred conclusion in line 2), but that she no longer subscribes to that (Mortensen 2012:82). Furthermore, the verb see is in the present tense. Th is is an example of dramatic present tense (Searle 2001:106) and might indicate that the utterance refers to a here and now in the interaction of both the training activity and the interview activity.
In combination with the relatively strong epistemic support from actually, the inference derived in relation to the training activity could be that the participant is now more certain about the instruction and her role as a mock suspect. As a contribution to the interaction in the interview activity, the use of the present tense might be a way of making her explanation more trust- worthy.
In summary, excerpt 1 shows two examples of how the interviewer reconstructs and emphasises the particular part of the interviewee’s utterance that signals certainty – or the lack of it. In both cases, the reconstruction evaluates the epistemic stance taking, thus, by itself forming an indirect question about trustworthiness and truth value. Th e interviewer expresses this question more explicitly in the requests for elaboration that follow.
Referring to the model for argumentative dialogue, another possible analysis could be that the reconstruction in line 3, holding the epistemic stance marker probably, is a clarifying warning about the following criticism or evaluation.
Nevertheless, I fi nd it reasonable to say that the reconstruction in itself holds both a reconstructive and a criticising and evaluating aspect, and that the interviewer makes this explicit in the following wh-question because the interviewee does not respond to the implicit questioning at fi rst.
Whether the interviewee’s utterances in excerpt 1 express a combination of the participant’s trouble remembering and the mock suspect’s unwillingness to talk to the police is not possible to tell for certain. It is possible to derive meaning in relation to the inferential frame of both the training activity and the interview activity, while not knowing which implicature is actually
intended by the interviewee. Th e interviewer seems to respond to the utterances as contributions to the interview activity, but his reconstructions of the ambiguous utterances fi t the inferential frames of both the training activity and the interview activity, thus questioning his co-participant’s memory and the mock suspect’s unwillingness respectively.
Later in the interview, an exchange similar to the one in excerpt 1 takes places. Prior to the exchange in excerpt 2 below, the interviewee has explained that during the quarrel Louai tries to stab the interviewee, and that Kim, the interviewee’s friend, tries to intervene. Th e interviewer asks the interviewee to elaborate on what was going on.
Excerpt 2
1 IE: yes (sighs), er, (p) I do not quite know how he grabs it I just think he is trying
2 to help me like in this situation because he knows that it can end up badly (.),
3 → er, (p) and then there’s this boy that that the Danish boy I think he had a knife
4 → too actually (p) 5 IN: → think
6 IE: well yes [it]
7 IN: [yes]
8 IE: → it he had a knife too
9 IN: → try to explain a bit about that
In the excerpt above, the interviewee says that she tror ‘thinks’ that the Danish boy had a knife too (line 3). By using the epistemic verb think, the interviewee marks her utterance as a subjective speech act with relatively weak evidential support (Mortensen 2012:82). However, she ends her statement with faktisk
‘actually’ in the fi nal position, which marks a stronger evidential support to the statement. By repeating think (line 5), the interviewer evaluates the epistemic basis for the statement. Th us, in the context this reconstruction functions as an elliptical declarative (Heinemann 2010:2705) that raises doubts about the subjective and epistemic part of the interviewee’s statement. Th e interviewee responds rather vaguely, though maintaining that the Danish boy had a knife (line 9). In her response she uses the past tense of the verb had. Compared to her use of the present tense in excerpt 1 (line 14), this could indicate that the utterance in this case functions as a contribution to the interaction in the interview activity, and that the intended inference is that the mock suspect is now standing by her explanation.
In excerpt 1 and 2, I have shown how the interviewer reconstructs those parts of the interviewee’s utterances that mark epistemic stance taking, thus questioning the truth value or certainty of the interviewee’s proposition, depending on which inferential frame the meaning is derived from. In excerpt 1, this brings the interviewee to revise her explanation, possibly because the expression of weak epistemic support actually derives from the participant’s trouble remembering her instruction, and the epistemic stance marker functions across the training activity and the interview activity. In excerpt 2, on the other hand, the interviewee maintains her explanation, presumably in this case staying in her role as a suspect.
Like in excerpt 1, the interviewee employs probably in a later utterance, shown in excerpt 3 below. Similar to excerpt 1 and 2, the interviewer reconstructs the interviewee’s utterance, but in this case he does not include the epistemic stance marker in the reconstruction. Th is could indicate a diff erent function of the reconstruction in this particular exchange.
To understand what is going on in excerpt 3, a brief summary of the interaction at the beginning of the interview might be helpful. In the preliminary free account, the interviewee explains that she was going to meet a friend at the day of the event under investigation. She says that it was “just a friend” and that she does not remember who he was – but it was not Louai.
A little later, the interviewee says that she does not feel like telling the friend’s name, again saying that it was “just somebody”. Obviously, this seems like the interviewee playing her role as the unwilling suspect Sedat trying to cover up the truth. However, in the light of the interviewee’s trouble remembering her instructions and the details of the mock case, an alternative interpretation could be that she is trying to make this uncertainty part of the role as an unwilling suspect. As already mentioned, the interviewee has trouble remembering the characters and the sequence of events in the mock case as it is described in the instructions. Th is, in particular, becomes clear and is said aloud by the interviewee right before the exchange in excerpt 3. Th e interviewee explicitly steps out of her role as a suspect and tells her co- participants that it is hard for her to understand and remember the instructions. Furthermore, she is confused about the character she is supposed to play in the ongoing interview, because she keeps mixing up the instructions and characters from the various training interviews she is supposed to take part in during the training course. Th is talk also reveals that she has made a wrong statement during the ongoing interview, and her co-participants, the interviewer and the observer, agree that she seems quite confused. Aft er this short exchange between the co- participants, the interviewer, without further notice, returns to the training interview. By reference to a printout from the interviewee’s mobile phone, he brings up the still unresolved issue of whether the interviewee was going to meet Louai or another friend, and whether the meeting between the interviewee and Louai was planned or not.
Excerpt 3
1 IN: but when I look at this report (.) secured phone (.) that is jo
your phone
2 IE: yes
3 IN: here I can see that there is this phone conversation on May 11
between you
4 and Louai (p) er both in the afternoon and in the evening (p) do you remember
5 what you talked about on the phone that day (p) I do realise that it is one day
6 ago but er it was yesterday
7 IE: → well it is jo probably because it was him then we were going
to meet jo
8 IN: → it was (.) Louai you were going to meet 9 IE: yes
10 IN: yes (p) okay so you arranged that with Louai 11 IE: well yes
12 IN: did you plan where you were going to meet
13 IE: well (p) we think supposedly we talked about meeting by the courtyard
In the excerpt above, the interviewee employs the epistemic stance marker nok ‘probably’ in her utterance (line 7). On the one hand, probably might refer to the causal relation between the interviewee’s and Louai’s phone call and the fact that they met at day of the event under investigation, that is, the reason for her conversation with Louai is that she did plan to meet him the next day.
On the other hand, probably might refer to the fact that it was Louai, and not another friend, the interviewee was going to meet. Each interpretation might be relevant, as both issues have been sources of confl ict and confusion during various stages of the interview.
Th e interviewee also employs jo ‘as you know’ twice in her utterance;
in relation to probably and in the fi nal position of her proposition. Seeing that probably does not per se specify the epistemic strength or the degree of probability (Hansen and Heltoft 2019:1058), but covers a spectrum “from the wildest guess to the most deliberate and watertight inference” ( Jacobsen 1992:15), a possible interpretation is that the interviewer responds to the interviewee’s use of probably as a statement of fact rather than conjecture.
Th is due to the strong epistemic support from jo. Th us, it seems that the interviewer does not use the reconstruction to raise doubt about the interviewee’s intentions or certainty, but to conclude and seek affi rmation that the interviewee had actually planned to meet Louai on the day of the event. In this case, the reconstruction seems to count primarily as a contribution to the mock interview and the interviewee’s utterance as a speech act with relatively strong epistemic support.
An alternative interpretation is that the interviewee’s utterance (line 7) is directed primarily at the training activity, and, within that inferential frame, functions as a question seeking affi rmation. As mentioned, jo might indicate strong epistemic support for the interviewee’s statement. Right before the exchange in excerpt 3, however, the interviewee has expressed anything but strong support for her explanation when she was talking with her co- participants about her confusion. Hence, in this instance, rather than indicating consensus by using jo twice, it seems that the interviewee is violating the maxim of quantity, thereby actually seeking confi rmation of the content of her proposition – and an affi rmation from her co-participant that she remembers her instruction correctly. In this perspective, jo in the fi nal position seems to function as an adverbial tag, thus turning the declarative into a question (Heinemann 2010: 2707). Jo is not commonly used as a tag, but due to its polyphonic instruction of consensus, and because particles like jo acquire dialogical meaning depending on their particular sentence position (Mortensen and Mortensen 2017:211, see also Jensen 2000, Christensen 2007), an interpretation of jo equalling a positive adverbial tag seems plausible.
Th e interpretation of the interviewee’s utterance as a question is further- more based on the cleft sentence it was him then we were going to meet jo (line 7). By employing this, the interviewee is both presupposing that she was going to meet somebody, and at the same time points out that, among other possible persons, this someone was Louai. Th is might refl ect the interviewee’s recurring confusion between the characters in the mock event, and the use of probably might, within the inferential frame of the training activity, indicate that this is what she believes to be the right interpretation of the instruction and the events in the mock case.
In other sequences during the interview, when the interviewee shows signs of a bad memory and confusion regarding her character, the interviewer seems to stay in his role as a detective and responds to the interviewee’s utterances as if they count as contributions to the activity of the mock interview. Th is is
also the case in excerpt 3, but the omission of the epistemic stance marker, probably, makes the reconstruction in this excerpt diff erent from the others.
Th is could indicate that the reconstruction is a contribution to the interaction between the participants in the training activity, and that the interviewer is responding to the interviewee’s utterance as a question, confi rming that she, his co-participant, remembers her instruction and role as a suspect correctly.
Th e interviewer does not explicitly step out of his role as a detective, though, but by this little change in the reconstruction, he seems to point out the relevant inferential frame and show that the training interview is back on track.
Another argument for interpreting the reconstruction as primarily a contribution to the training activity is the questioning function of the interviewee’s utterance prior to the reconstruction. An exchange like that, where the suspect asks questions about the event under investigation, does not fi t the conventional form of an investigative police interview, including the casting of the detective as the questioner. Hence, in the light of the constraints on allowable contributions in the two diff erent activity types and the preceding exchange between the co-participants, it seems plausible that the inference is to be derived primarily from the inferential frame of the training activity.
7. Summarising the interview – and the analysis
When the interview reaches the closing stage, the interviewer summarises what the interviewee has told about the mock event under investigation, and he lets the interviewee know that he does not believe her. For instance he says that part of the interviewees’ explanation ”sounds a little odd” and places the responsibility for this oddness and the lack of information on the interviewee by saying that he will write in his report that the interviewee refuses to make suffi cient statements about the confl ictual issues. Th us, he indicates that he believes that the unclarifi ed parts of the event are the result of the interviewee’s unwillingness to tell rather than a bad memory.
As shown in excerpt 1 and 2, the interviewer raises doubts about the uncertainty and trustworthiness of the interviewee during the questioning stages of the interview. In these examples, however, he mitigates the confrontation in the follow up questions, saying try to tell a little more about that (excerpt 1, line 9) and yes why are you saying that it is probably him who had it can you explain a bit about that (excerpt 1, line 5-6). In the closing summary, the interviewer does not mitigate his utterances like this, and it seems that during the interview he has compensated for the potential face threats that his questioning poses to the interviewee. Th is may be due to the interviewer’s interpretation of the weak epistemic support to the propositions as an indication of the co- participant’s uncertainty rather than the mock suspect’s unwillingness. Th at is, considering the propositions as, maybe intended, implicata within the inferential frame of the training activity. Consequently, his reconstructions have fulfi lled a double function
and operated across the training activity and the interview activity by raising doubts about the trustworthiness of the mock suspect while also responding to the derived implicatum and supporting the co-participant. In the context of the training activity, the reconstructions of the interviewee’s utterances are not questioning the truth value or the trustworthiness of the interviewee, but evaluate the consistency between the instruction and the interviewee’s statement. Whether or not such double function can explain why the interviewer mitigates his utterances and compensates for the face threat of a co-participant, is not obvious. Nevertheless, the interviewee shows more certainty and unwillingness to cooperate when she is clearly enacting the role identity of the suspect at the end of the interview. Hence, the interviewer does not have to protect a co-participant’s face, and in that situation, he seems less concerned with mitigating his utterances when he questions the interviewee’s intentions.
8. Conclusions
Th e interview analysed in this paper is a mock police interview conducted as part of a training course for skilled police offi cers. In the training interview, the participants navigate two distinctly diff erent activity types, developing within diff erent inferential frames; the training activity and the interview activity.
Th e interview activity simulates an asymmetric relation between the participants who are instructed to act as interlocutors of an interaction that does not take its natural point of departure in mutual expectations of social cooperativeness. In the training activity, though, the participants are part of a symmetric relation, and they share the same goals for the communication, among these to get the most out of the training interview and keep it on the right track. Th is activity invites to social cooperation. In the analysis, I have shown how contributions to the interaction function across these two activities, even though the inferential frames and the conditions for these activities are inherently diverse.
In the excerpts displayed in this paper, both the socially non-c ooperative and cooperative aspects show, because one of the participants struggles with remembering her instruction. When her utterances seem to express this struggle and uncertainty as a participant in the training activity, the interviewer makes his contributions to the interaction interpretable in relation to both the training activity and the interview activity. Th e analysis illustrates how the interviewer by reconstructive speech acts manages to evaluate the interviewee’s utterances across the activities. With the reconstructions, the interviewer raises doubt about the trustworthiness of the mock suspect, hence contributing to the role-played interaction, and simultaneously addresses the uncertainty of the co-participant, thereby supporting the participant and keeping the training activity on the right track.
By applying Nielsen’s model of argumentative dialogue to the training interview, I have illustrated how the model has the potential to capture central features of confl icting dialogue within diff erent activity types. In particular, the speech act types reconstruction and criticism are essential in the training interview. In this context, the reconstructions function as questions that, within the inferential frame of this particular activity type, oft en serve both a clarifying and a criticising function.
Th e analysis also shows how the participants negotiate comprehension and possible inferences. Whether the derived implicatum is intended by the interviewee, is not possible to tell from the excerpts, as her propositions in most cases can count as statements from both the mock suspect and the participant in the training course. Th e interviewer responds to this double possibility of inference, and apparently forms his reconstruction as c ontributions to the interview activity, while also making an interpretation within the inferential frame of the training activity possible.
Th ese contributions, holding diverse possible implicata, show how the participants take part in the reciprocal act of implicature, and how the activity- crossing implicatures both contribute to the mock interview and simultaneously support the participants and guide the training interview.
Appendix 1, excerpt 1 (Danish)
1 IN: hvor skulle den kniv i så fald være kommet fra
2 IE: → (p) jamen altså det jo (p) det er jo nok ham der Louaia der der havde den (p)
3 IN: → nok ham der Louaia der havde den 4 IE: ja
5 IN: → ja (p) hvorfor siger du det nok er ham der havde den (.) kan du forklare lidt om
6 det
7 IE: → ej men altså det er jo fordi da da han (p) da han begyndte og og skælde mig ud
8 der ikke øh 9 IN: ja
10 IE: → om øh min familie så så øh så blev jeg jo rigtig sur jo og sådan noget og så
11 begyndte jeg jo også og skubbe lidt til ham (.) 12 IN: okay
13 IE: til til Louaia der så så så derfor så øh så så lige pludselig så
blev jeg bange
14 fordi så kunne jeg lige pludselig se en at: at at Louaia havde en kniv på sig
15 IN: → du siger du ser en kniv så 16 IE: ja
17 IN: ja
18 IE: → jeg ser [faktisk en kniv]
19 IN: → [prøv at fortælle] lidt mere om det hvis du prøver og 20 IE: ja
21 IN: genopfrisk øh hvornår ser du den kniv første gang
Appendix 2, excerpt 2 (Danish)
1 IE: Ja (sukker lidt) øhm (p) jeg ved ikke helt hvordan han tager den jeg tror bare
2 han prøver på og hjælpe mig ligesom i den her situation fordi han godt ved at
3 det kan gå galt (.) øhm (p) og så er der den her dreng der den der danske dreng
4 → jeg tror også han havde en kniv faktisk (p) 5 IN: → tror
6 IE: jaer (.) [det]
7 IN: [ja]
8 IE: det han havde også en kniv 9 IN: → prøv at forklare lidt om det
Appendix 3, excerpt 3 (Danish)
1 IN: men når jeg kigger på den her rapport sikret telefon (.) som jo er din telefon
2 IE: ja
3 IN: der kan jeg se der er den her telefonsamtale 11. maj mellem dig og Louai (p)
4 øhm det er både om eftermiddagen og om aftenen (p) kan du huske hvad du
5 snakkede med ham om i telefonen den dag (p) jeg ved godt
det er en dag siden 6 men øh det var i [går]
7 IE: → [jamen] det er jo nok fordi det så var ham vi skulle mødes
med jo
8 IN: → det var (.) Louai I skulle mødes med 9 IE: ja
10 IN: ja (p) okay så det aftalte du med Louai 11 IE: jaer
12 IN: aftalte I hvor I skulle mødes henne
13 IE: jaer (p) vi tror vist nok vi snakkede om vi skulle mødes ved gården
Notes
* I thank the anonymous reviewer and my editor Sune Sønderberg Mortensen for insightful comments and for pointing out crucial points of improvement. In particular, I have benefi tted from Sune Sønderberg Mortensen’s great involvement in fruitful and encouraging input and discussions.
I also owe thanks to Tanya Karoli Christensen, who put me in touch with Regin Hviid and provided me with the necessary audio fi les. I am especially grateful to Regin Hviid for allowing access to these fi les and the huge body of data from his study and, not least, for his willingness to share his knowledge about investigative interviewing and training within the Danish Police.
1 PEACE: Planning and preparation, Engage and explain, Account, Closure, And Evaluation of the interview and the interviewer’s performance (Clarke and Milne 2001:1)
2 For describing and analysing this kind of communication, Linell and Th unqvist suggest the dynamic concept of framing rather than frame (see Linell and Th unqvist 2003).
3 I refer to the participants as interviewer and interviewee. Th e latter is the participant who plays the role as a suspect, and the interviewer is the detective who is doing the interview.
4 Archer (2005:58) distinguishes between Gricean conversational cooperation and social cooperation.
5 Note that Nielsen uses ‘speaker’ and ‘hearer’ instead of the argumentative roles.
Th e reason is that “while the speaker of a reconstructive or critical speech act is necessarily identical to the argumentative role of antagonist, it is not given that the hearer is necessarily identical to the protagonist role” (Nielsen 2005:206).
6 Th e comments about the felicity conditions for reconstructions and criticism are Nielsen’s own.
7 Th e description of the event under investigation is based on the instruction material from the experimental study (Hviid 2019:99).
8 Due to lack of space the quantitative overview is not presented in this paper.
9 Th e excerpts are English translations, except for the Danish dialogical particle jo (‘as you know’) as it has no direct translation equivalent in English. Th e original Danish excerpts are to be found in appendices 1, 2, and 3.
10 IN: the interviewer. IE: the interviewee.
References
Archer, Dawn. 2020, this issue. Understanding questions and answers in context : An argument for multi-channel analysis. Scandinavian Studies in Language 11(1), 196-213.
Archer, Dawn. 2005. Questions and answers in the English courtroom (1640–1760): A sociopragmatic analysis. Amsterdam: John Benjamins Publishing Company.
Bitsch, Anders and Erik Modin. 2016. Politiarbejdets grundmodel. Adam Diderichsen and Anne-Stina Sørensen (eds.) Den skarpe ende. Grundbog i politiarbejde.
Frederiksberg: Samfundslitteratur.
Christensen, Tanya K. 2007. Hyperparadigmer – en undersøgelse af paradigmatiske samspil i danske modussystemer, PhD thesis. Roskilde: Roskilde University.
Clarke, Colin and Rebecca Milne. 2001. National evaluation of the PEACE investigative interviewing course. PRAS 149. London: Home Offi ce.
Eemeren, Franz H. van and Rob Grootendorst. 1992. Argumentation, Communication, and Fallacies. Hillsdale, New Jersey: Lawrence Erlbaum.
Eemeren, Franz H. van and Grootendorst, Rob. 1984. Speech Acts in Argumentative Discussions. Dordrecht: Foris Publ.
Ennis, Robert H. 1982. Identifying Implicit Assumptions. Synthese 51(1), 61–86.
Fisher, Ronald P. and R. Edward Geiselman. 2010. Th e Cognitive Interview method of conducting police interviews: Eliciting extensive information and promoting Th erapeutic Jurisprudence. International Journal of Law and Psychiatry 33, 321–
328.
Gibbons, John P. 2003. Forensic Linguistics. An Introduction to Language in the Justice System. Blackwell Publishing.
Grice, Herbert P. 1975. Logic and Conversation. P. Cole and J. L. Morgan (eds.) Syntax and Semantics (3), Speech Acts. New York: Academic Press 1975, 41–58.
Grootendorst, Rob, Ralph H. Johnson, Christian Plantin, Charles A. Willard, Frans H. van Eemeren and Francisca Snoeck Henkemans. 1996. Fundamentals of argumentation theory : A handbook of historical backgrounds and contemporary developments. Mahwah: Taylor & Francis Group.
Hansen, Erik and Heltoft , Lars. 2019. Grammatik over det Danske Sprog. Copenhagen / Odense: Det Danske Sprog- og Litteraturselskab / Syddansk Universitetsforlag.
Harris, Sandra. 2011. Interrogation and evidence : Questioning sequences in courtroom discourse and police interviews. Christopher N. Candlin and S. Sarangi (eds.) Handbook of Communication in Organisations and Professions. Berlin: De Gruyter.
Heinemann, Trine. 2010. Th e question-response system of Danish. Journal of Pragmatics 42(10), 2703–2725
Hviid, Regin E. 2017. Information Yield from Investigative Interviewing : A Randomised Laboratory Trial of Simultaneous vs. Post-Interview Report Writing. Cambridge:
University of Cambridge.
Jacobsen, Henrik G. 1992. Vist og nok: Om et par formodningsbiord i dansk. Mål og mæle 15(2), 13–20.
Jensen, Eva S. 2000. Danske sætningsadverbialer og topologi i diakron belysning, PhD thesis. Copenhagen: University of Copenhagen.
Levinson, Stephen C. 1992. Activity types and language. Paul Drew and John Heritage (eds.) Talk at work. Interaction in institutional settings : Studies in interactional sociolinguistics 8. Cambridge: Cambridge University Press, 71–97.
Linell, Per and Daniel P. Th unqvist. 2003. Moving in and out of framings: activity contexts in talks with young unemployed people within a training project. Journal of Pragmatics 35(2003), 409–434.
Martin, Jim R. and Peter R. R. White. 2005. Th e Language of Evaluation: Appraisal in English. Basingstoke: Palgrave Macmillan.
Mortensen, Janus. 2012. Epistemisk positionering i dansk talesprog. NyS – Nydanske Sprogstudier 42. Dansk talesprog 2012, 62–91.
Mortensen, Sune S. 2020, this issue. A question of control? Forms and functions of courtroom questioning in two diff erent adversarial trial systems. Scandinavian Studies in Language 11(1), 239-278.
Mortensen, Sune S. 2018. ”Det tror du ikke på?” Deklarative spørgsmål som grammatikaliserede metaspørgsmål. Ny forskning i grammatik 25, 107–129.
Mortensen, Sune S. and Janus Mortensen. 2017. Epistemic stance in courtroom interaction. Poggi, F. and A. Capone (eds.) Pragmatics and law : Practical and theoretical perspectives. Springer International Publishing, 401–437.
Nielsen, Niels M. 2005. Counter Argument – In Defence of Common Sense. Roskilde:
Roskilde University.
Nielsen, Niels M. 2010. Argumenter i kontekst. Introduktion til pragmatisk argumentationsanalyse. Frederiksberg: Samfundslitteratur.
Noh, Eun-Ju. 1998. Echo Questions: Metarepresentation and Pragmatic Enrichment.
Linguistics and Philosophy 21(6), 603–628.
Searle, John. 2001. How Performatives Work. Vanderveken, Daniel, Kubo, Susumu and Vanderkeven, Daniel (eds.) Essays in speech act theory. Pragmatics and Beyond New Series 77. Amsterdam: John Benjamins, 85-107.
Searle, John. 1976. A Classifi cation of Illocutionary Acts. Language in Society 5(1), 1–23.
Stivers, Tanya and Nick J. Enfi eld. 2010. A coding scheme for question–response sequences in conversation. Journal of Pragmatics 42(10), 2620–2626.
Teleman, Ulf, Staff an Hellberg and Erik Andersson. 1999. Svenska Akademiens grammatik 4: Satser och meningar. Stockholm: Svenska Akademien.
Vrij, Aldert. 2008. Detecting Lies and Deceit: Psychology of Lying and its Implications for Professional Practice. Second edition. Chichester: Wiley.
White, Peter R. R. 2003. Beyond modality and hedging : A dialogic view of the language of intersubjective stance. Text & Talk 23(2), 259–284.
